1
$\begingroup$

If $Y$ has a binomial distribution with $n$ trials and probability of success $p,$ show that the moment-generating function for $Y$ is $m(t) = (pe^t + q)^n,$ where $q = 1 − p.$

I got to $$m(t) = E(e^{tY}) = \sum_{y=0}^n \binom{n}{y}(pe^t)^yq^{n-y}.$$ How to go from here?

$\endgroup$
2
$\begingroup$

As Y is a discrete random variable, the moment generating function can be computed quite easily. Your start is good. Now, remember that the sum over all possible binomial coefficients on $N$ can be simplified: \begin{eqnarray} M(t) = \mathbb{E}[e^{tY}] &=& \sum_{n= 0}^N e^{tn} {N \choose n} p^n q^{N-n} \\ &=& \sum_{n = 0}^N \left(pe^t\right)^n {N\choose n} q^{N-n}\\ &=& \left(pe^t + q\right)^N \end{eqnarray}

$\endgroup$
2
$\begingroup$

Another way is noting that if $X \sim \text{Bin}(n,p)$ then $X \stackrel{d}{=} \sum_{k=1}^n Y_k$ if $Y_1,...,Y_k$ are i.i.d. Bernoulli distributed with succes parameter p. And you can easily show that $M_{Y_1}(t)= 1-p+p\exp(t)$. Thus $$ \begin{align} M_X(t) &= E[e^{tX}]\\ &= E\left[ \exp\left(t \sum_{k=1}^n Y_k \right)\right] \\ &= E\left[ \prod_{k=1}^n\exp\left(t Y_k \right)\right] \\ &= \prod_{k=1}^nE\left[\exp\left(t Y_k \right)\right] \\ &= \prod_{k=1}^n (1-p+p\exp(t)) \\ &=(1-p+p\exp(t))^n \end{align} $$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.